Overview
A Model Context Protocol (MCP) server for Prometheus. This MCP server provides access to Prometheus metrics and PromQL queries through standardized MCP interfaces, enabling AI assistants to execute PromQL queries and analyze metrics data. It supports features such as executing PromQL queries, discovering and exploring metrics (listing metrics, retrieving metadata, and viewing instant and range query results), authentication via basic credentials or bearer tokens, and Docker containerization for deployment. Configuration is done via environment variables (PROMETHEUS_URL, PROMETHEUS_USERNAME, PROMETHEUS_PASSWORD, PROMETHEUS_TOKEN, etc.). Transport modes include stdio, HTTP, and SSE, with configurable bind host and port, and optional custom headers. The MCP server is designed to be used with MCP-compatible clients (Claude Desktop, Claude Code, VS Code, Cursor, Windsurf, etc.). The README also outlines installation and deployment options (Docker, Docker Desktop/MCP Toolkit, Claude setup) and lists a suite of discovery and query tools to interact with Prometheus data.
Features
Execute PromQL queries against Prometheus
Run PromQL queries against a Prometheus server through MCP interfaces.
Discover and explore metrics
Find and inspect available metrics and their metadata via MCP.
List available metrics
Provide a catalog of metrics exposed by Prometheus.
Get metadata for specific metrics
Retrieve metadata for individual metrics to aid analysis.
View instant query results
Return instant PromQL query results.
View range query results with steps
Provide range query results with adjustable step intervals.
Authentication support (env-based)
Support Basic auth and Bearer token authentication via environment variables.
Docker containerization support
Containerized deployment via Docker with configurable environment variables.
Who Is This For?
- AI assistants:Interact with Prometheus via MCP to run queries and analyze metrics.
- Developers:Integrate Prometheus data into MCP-enabled tools and workflows.
- DevOps teams:Access metrics through MCP in DevOps workflows.
